Some of the unsuccessful attempts to produce voluntary alcohol dependence in rats have involved placing the animals under stressful conditions. However, a review of the literature of experimental neurosis reveals that the stressors employed are not those leading to behavioral or physological pathology in these animals. We propose to examine alcohol consumption in terms of the above criteria under conditions which do lead to compulsive maladaptive behavior or other pathological symptoms in rats. These are: (1) brief, intense, unavoidable shock (Pre-shock) coupled with either (2) prolonged (30 day) exposure to conflict or (3) an insoluble problem which forces the animal to repeatedly choose between two potentially punishing events (avoidance-avoidance conflict). There is a strong suggestion (Maier, 1949) that in rats, once the pathological process is initiated, the choice of symptom will be determined by simple availability of a given response, rather than any intrinsic appropriateness. We therefore suggest that by pretraining rats to drink alcohol, and making such solutions available in the above situations, alcohol dependence may replace or supplement the symptoms normally observed.